Prediction by measuring neutron porosity by using artificial intelligence for ATN-103 well in Atallah North Field

Abstract

Our study, through the application of artificial neural networks technology, aimed at obtaining the NPHI measurement at the ATN-103 well, whereby the implemented neutron measurement had been subjected to a defect that did not show correct results, the network input parameters and its output target were determined, and a rule was divided The data for the wells adjacent to a training well, through which the network that we implemented was trained on the IP-V3.5 interpretation program to obtain the optimal network structure in obtaining the most accurate results, and through two wells the performance of the trained network was tested, and the results of this network were compared with the measurements The neutrinos executed in them, the difference between the two values ​​representing the values ​​of the prediction error was calculated, and finally the network was generalized in the ATN-103 well and the results were highly accurate (95%).

During the study, the formations were evaluated qualitatively using different cross-plots to determine the predominant lithological pattern in the study area and porosity in general, and the proportion of each of them, as we have identified the pattern of clay are placed in components in which the increase in the clay volume based on  Thomas-Stepper. We have calculated by quantitative evaluation, the clay volume using Gama-Ray log and the total and effective, primary and secondary porosity, as well as water saturation and total water volume in the clean zone and flashed zone.